基于多智能体的大规模人群疏散模拟技术研究-计算机应用技术专业论文.docxVIP

武汉理工大学硕士学位论文摘要 武汉理工大学硕士学位论文 摘要 随着三维动画技术的发展以及计算机软硬件能力的提高，虚拟现实技术在 人们的实际生活中发挥着越来越重要的作用，而对群体仿真技术的研究也在公 共安全领域起到了较为突出的作用和影响。 本文研究了基于多智能体的大规模人群疏散模拟技术，并建立实现了一个 群体疏散系统，其主要功能模块包括：目标驱动智能体设计、群体路径规划算 法设计、人群和场景建模以及实时渲染。 本文的研究重点是目标驱动智能体设计和群体路径规划算法设计。目标驱 动的智能体是将环境中的每个人都设计成为一个智能体，智能体具有自主感知 环境，同时进行自主决策的能力，它的行为是受目标驱动的。 另外，本文介绍了智能体的路径搜索算法，如Dijkstra算法和A宰寻径算法。 为了使智能体更有效地感知到自身的环境状况，从而提高路径搜索效率，本文 引入并设计了运动导航机制，包括可视点导航图和导航网格，并在这两种导航 网格的基础之上，提出了一种基于洪水填充算法的层次导航网格，包括宏观网 格和微观网格。 最后，本文对于系统中的人群和环境建模进行了简单介绍，包括个性化的 人群建模方法和三维场景建模、场景管理技术。最后对模拟过程的实时渲染技 术进行了简单讲解，并引入真实案例证明了该系统的有效性和实用性。 关键词： 多智能体、人群疏散、路径规划、导航、大规模 武汉理工大学硕士学位论文Abstract 武汉理工大学硕士学位论文 Abstract As the development of 3D animation technologies and computer hardware and software capabilities，the technology of virtual reality is playing a more and more important role in our real life．At the same time，the researches on crowd simulation technique have also made some salient influences in public safety areas． In this paper,we have made some researches on large—scale crowd evacuation techniques based on multi-agent，and established a crowd evacuation system．The main function modules of this system includes as following：goal-driven agent design, path planning algorithm design，crowd and scenario modeling and real—time rendering． The focuses of this paper are the first two parts：goal-driven agent design and path planning algorithm design．Goal—driven agent is to regard every individual in the environment as an agent，it can perceive its own environment，make decision and take some behaviors independently,its behaviors are driven by its goals． Besides，this paper introduced some path-searching algorithms，such as Dijkstra algorithm and A木algorithm．In order to improve the path-searching efficiency,we introduced and designed a mechanism called navigation graph，including view navigation map and navigation grid．And then，we proposed a multi-level navigation gad based on flood filling algorithm． Finally,the Paper made a brief introduction of crowd modeling and 3D sce